no code implementations • 20 Jun 2023 • Ammar Abbas, Sri Karlapati, Bastian Schnell, Penny Karanasou, Marcel Granero Moya, Amith Nagaraj, Ayman Boustati, Nicole Peinelt, Alexis Moinet, Thomas Drugman
We show that eCat statistically significantly reduces the gap in naturalness between CopyCat2 and human recordings by an average of 46. 7% across 2 languages, 3 locales, and 7 speakers, along with better target-speaker similarity in FPT.
no code implementations • 27 Oct 2021 • Ayman Boustati, Hana Chockler, Daniel C. McNamee
In this study, we apply causal reasoning in the offline reinforcement learning setting to transfer a learned policy to new environments.
no code implementations • NeurIPS 2020 • Ayman Boustati, Omer Deniz Akyildiz, Theodoros Damoulas, Adam Johansen
We introduce a framework for inference in general state-space hidden Markov models (HMMs) under likelihood misspecification.
1 code implementation • NeurIPS 2020 • Lorenz Richter, Ayman Boustati, Nikolas Nüsken, Francisco J. R. Ruiz, Ömer Deniz Akyildiz
We analyse the properties of an unbiased gradient estimator of the ELBO for variational inference, based on the score function method with leave-one-out control variates.
no code implementations • 9 Mar 2020 • Ayman Boustati, Sattar Vakili, James Hensman, ST John
Approximate inference in complex probabilistic models such as deep Gaussian processes requires the optimisation of doubly stochastic objective functions.
no code implementations • 23 Feb 2020 • Ayman Boustati, Ömer Deniz Akyildiz, Theodoros Damoulas, Adam M. Johansen
We introduce a framework for inference in general state-space hidden Markov models (HMMs) under likelihood misspecification.
no code implementations • 29 May 2019 • Ayman Boustati, Theodoros Damoulas, Richard S. Savage
We present a multi-task learning formulation for Deep Gaussian processes (DGPs), through non-linear mixtures of latent processes.